01 First things first

Product Lifecycle Management (PLM)

Product Lifecycle Management (PLM) is a strategic approach to developing, managing, and improving products from conception to disposal—a way of dealing with the different stages across a product lifecycle. However, it can also be a piece of software (or system) that helps manufacturing organizations and Engineering-to-Order (ETO) companies efficiently work through these different stages.

By blending existing procedures and processes with individual expertise and innovative technology, PLM software like Siemens Teamcenter provides a framework that enhances product quality, reduces costs, and accelerates time to market. Product Lifecycle Management software offers a single platform for all product data and related processes. This single source of truth makes it easier for stakeholders to find the most up-to-date information, allowing them to make the right decisions more quickly and efficiently.

02 The stages of PLM

What, when, and why?

From a manufacturing and ETO perspective, Product Lifecycle Management can be divided into five main stages: Conception, Design and Engineering, Manufacturing, Commissioning, and Decommissioning.

{{second-first}}

‍

{{second-second}}

‍

{{second-third}}

‍

{{second-fourth}}

‍

{{second-fifth}}

‍

03 The benefits of PLM

How can PLM help?

The benefits of Product Lifecycle Management for manufacturing aren’t just linked to transparency and timekeeping. Clear protocols facilitated by comprehensive PLM software like Siemens Teamcenter increase the likelihood of creating better-quality products, fewer errors, and greater cost savings thanks to more efficient production processes.

In short, PLM software is crucial for both custom ETO requests and mass-produced products.

‍

{{third-first}}

‍

{{third-second}}

‍

{{third-third}}

‍

{{third-fourth}}

‍

{{third-fifth}}

‍

04 The key components of PLM software

Optimizing the PLM value chain

PLM software streamlines the way different manufacturing companies and specific stakeholders can access data. This is done by integrating tools and features to optimize the overall management of a product. Some tools, such as CAD software, are used heavily at specific stages, whereas key components like document management make up the backbone of a PLM system’s overall offering.

Siemens Teamcenter offers a multitude of tools and components that make PLM a no-brainer for manufacturers looking to scale and optimize their business processes without losing track of the original vision for the brand and products.

‍

{{fourth-first}}

‍

{{fourth-second}}

‍

{{fourth-third}}

‍

{{fourth-fourth}}

‍

{{fourth-fifth}}

‍

{{fourth-sixth}}

‍

{{fourth-eighth}}

‍

{{fourth-seventh}}

‍

05 Picking a PLM implementation partner

Ask yourself the right questions

Picking a PLM partner is the first step to increased efficiency, smoother processes, and better data management. However, to ensure your business's needs are met now and in the future, it's worth considering a few things.

‍

{{fifth-first}}

‍

{{fifth-second}}

‍

{{fifth-third}}

‍

{{fifth-fourth}}

‍

{{fifth-fifth}}

‍

{{fifth-sixth}}

‍

06 Digital transformation with CLEVR

Product Lifecycle Management in action

Siemens Teamcenter is a comprehensive PLM software suite offering extensive capabilities for managing product data and processes across the entire product lifecycle.

We chose to partner with Siemens because of Teamcenter’s collection of tools and integrations, as well as its overall usability.

Nel Hydrogen recently partnered with CLEVR to significantly enhance its product development capabilities. By leveraging Siemens Teamcenter, CLEVR is implementing a comprehensive PLM solution that streamlines data management and helps automate engineering processes. The collaboration is ongoing, with a view to expanding the scope of this initial project.

Our expertise in digital transformation and PLM is what sets us apart from other solution partners. We combine extensive industry knowledge with digitalization expertise to implement tailor-made Siemens Teamcenter solutions that automate and streamline product lifecycle processes.

Even as your company scales and adapts to new challenges, your processes remain flexible and robust. Let CLEVR guide you through today’s bold decisions for greater peace of mind.

Design and Engineering

This stage includes hands-on tasks that bring a concept to life; detailed product designs, specifications, and prototypes are the name of the game. Tools like CAD systems help designers visualize ideas, enabling engineers to create prototypes.

Quality Assurance and Engineering departments in larger manufacturing organizations use prototypes to ensure a product meets design and performance requirements before mass production. Feedback from testing highlights the refinements needed for validation.

ETO companies often use virtual prototypes, models, and simulations during this stage. Avoiding too many physical iterations helps keep costs low for businesses that can't benefit as much from economies of scale.

Conception

During the ideation phase, competitive analyses help identify market gaps and customers’ unserved needs. This information is used to conceptualize the product, creating a solid foundation for the subsequent PLM stages and decision-making processes.

Automotive manufacturers may, for instance, conduct a competitive analysis to identify gaps in the market for electric trucks, conceptualizing a new model that meets specific urban delivery service needs.

Manufacturing

From a mass manufacturing perspective, this stage starts with a validated, market-ready product resulting from iterative feedback rounds during development. Once the production process is established, it’s time to scale. Planning, executing, and monitoring the scaled production process involves supply chain management and quality control.

ETO companies usually have a single manufacturing process and only one chance to get an order right. Therefore, this stage depends heavily on accurate information from the Design and Engineering, facilitated by efficient PLM software that gets the right information to the right people at the right time.

Commissioning

For mass manufacturers, this stage consists mainly of introducing the product to the market, distribution, sales, and support. Successful product launches require these aspects to be aligned from the start.

In an ETO context, commissioning involves customizing a product's delivery, installation, and support. Successfully deploying bespoke products requires careful logistics coordination, detailed installation procedures, and tailored customer support.

Managing product effectivity—acquiring spare parts and documentation for a specific product version—is also crucial here.

PLM software helps manage these complex processes by providing precise, up-to-date information to all stakeholders. For example, in an ETO machinery project, PLM ensures that engineering details, installation guides, and support documentation are all aligned, allowing for a smooth transition from production to customer site setup and ongoing support.

Decommissioning

Product decommissioning involves Product Managers, Environmental Compliance personnel, and logistics teams. Retirement isn’t just stopping production—effective communication with customers and suppliers is crucial. A tech company may need to plan for disposing of, recycling, or remanufacturing obsolete laptops, ensuring the remaining stock is sold off or used for spare parts. Letting the right people know exactly how these processes should be expected to work is almost as important as the procedures themselves.

For ETO companies, decommissioning involves carefully planning the phase-out of custom products and ensuring clients are supported throughout the process.

Enhanced product quality

PLM software creates a single source of truth for all product data, giving (authorized) departments and stakeholders access to the latest information. This comprehensive data management reduces errors resulting from miscommunication or outdated information.

PLM software also supports extensive testing and validation processes, which helps manufacturers identify issues early in the development cycle.

Reduced time to market

PLM software streamlines a product’s development stage by automating workflows and improving communication among teams. Reducing the time spent on administration speeds up decision-making and helps avoid human errors often caused by repetitive, manual tasks.

Enhanced data management and collaboration also improve the efficiency of the earlier lifecycle stages, which leads to quicker market introductions.

Better data management and collaboration

A centralized PLM system ensures that all product data is easily accessible to those who need it, such as marketers creating assets or campaign messages and after-sales personnel creating training assets for customer support staff. This improves data accuracy and consistency, enabling more informed decision-making. PLM software allows and encourages departments to share information in real time, which reduces information silos and keeps everyone on the same page with the most up-to-date information. 

Cost savings across the product lifecycle

PLM software helps companies avoid inefficient practices that often clog up business processes. This helps reduce costs associated with product development, manufacturing, and maintenance. It also supports better resource management and reduces the need for costly reworks.  

An overview of the production process, including governance and control of automated machinery, lets companies spot material waste and identify ways to optimize production schedules. This reduces manufacturing costs linked to energy consumption and raw materials, which minimizes the environmental impact of a company’s operations. Siemens Teamcenter offers a Carbon Footprint Calculator to help companies assess their decisions as they look to strike a balance between environmental impact, cost reduction, and meeting customer demands. 

Integration and connectivity

Siemens Teamcenter offers extensive integration capabilities with real-time data access for better collaboration. This ensures that all departments and stakeholders across the product lifecycle are on the same page. This is crucial for ETO manufacturers and larger organizations aiming to streamline operations, maintain product quality, and scale effectively.

Good PLM software should seamlessly integrate with various enterprise systems and authoring tools, ensuring cohesive product data management throughout its lifecycle. This means creating a seamless flow of information by connecting Enterprise Resource Planning (ERP) systems, Computer-Aided Design (CAD) tools, and document management software.

Computer-aided design (CAD)

CAD software is essential for creating precise 2D and 3D models, allowing engineers and designers to visualize and iterate on product designs. In PLM, CAD integrates design data with other lifecycle processes, ensuring that all design changes are tracked and managed efficiently. As you’d imagine, CAD software is heavily involved in the conception stage of a product’s lifecycle. So is Product Data Management. 

Product Data Management (PDM)

PDM centralizes all product-related data—which often changes—ensuring accessibility, accuracy, and security. This invariably improves collaboration and decision-making. Within PLM, PDM manages the lifecycle of product data, including version control and access permissions, ensuring that the latest information is available to the right people. 

Bill of Materials (BOM)

A bill of materials (BOM) lists all materials, parts, and assembly configurations required to manufacture a product, which makes it a key feature of the development stage. A BOM represents the product structure in a hierarchical format that clearly presents the relationship between certain components and assemblies. Depending on the product and industry, a BOM can range from a simple, single-level structure to a multi-level structure with specific manufacturing, engineering, and customization guidance.

Like PDM systems, BOM systems track changes. This means that any requested changes to a BOM are documented and sent for approval. A BOM can also include tools to analyze the cost of materials and components. Having an exhaustive and holistic view of the costs will help manufacturers with budgeting forecasts, general cost management, and reporting.

Engineering change management

Engineering Change Management is the tracking, controlling, and approving of changes to product designs and processes. During the development stage, Engineering Change Management helps stakeholders assess the impact of proposed changes on existing designs and processes. It also records modifications, which is vital with the rapid development of a product often containing so many iterations—some of which may need to be revisited for another assessment. 

Computer-Aided Manufacturing (CAM)

CAM software automates manufacturing by converting CAD models into machine instructions, enhancing production precision and efficiency. In PLM software, CAM ensures that manufacturing data is consistent with design data, reducing errors and streamlining the transitions between the design, development, and production stages. 

Supply Chain Management (SCM)

SCM tools are used in the launch and production phase to manage the flow of goods, information, and finances related to a product. In PLM, SCM ensures that supply chain activities are aligned with product development and production schedules, which improves efficiency and reduces costs. 

Document management

This process comprises organizing and managing all documents related to a product’s entire lifecycle. This can include items ranging from compliance records to product brochures. Having the necessary documents in easy-to-find places is key when companies are posed with compliance questions from external regulators. This component is often a feature of the end-of-life phase when companies look to “close the loop” of an existing product, ensuring that it has been produced, distributed, and discontinued in a manner that complies with any number of (changing) regulations.

Compliance and regulatory management

Maintaining a database of the regulations and standards applicable to a product is critical for keeping stakeholders informed on the latest regulatory developments. Sudden changes can result in product non-compliance, which invariably leads to fines and can negatively impact publicity and trust. 

This key component provides the tools to track compliance throughout a product’s lifecycle, which helps generate reports needed for regulatory submissions. Audits can often be lengthy and nerve-wracking for companies. So, having an automated process in place to ensure products meet safety and quality standards can help avoid surprises when regulators are sifting through documentation. 

Do they provide an end-to-end solution?

Ensure the PLM partner you choose will handle the entire product lifecycle. Those that appear only at certain stages and offer support reactively may struggle to produce the most efficient results for your business.

Are they innovative?

It's good to consider how and if your potential PLM partner embraces new technology. Some tried-and-tested methods are all well and good, but partners that embrace the power of low-code with novel PLM systems like Siemens Teamcenter could provide the spark you need to bring your product processes to the next level.

Do they have the right expertise?

Verifying the expertise of those you're considering to partner with is crucial. How experienced are they when it comes to implementing PLM solutions? Do they have the right connections and partnerships with software providers?

Will they be the right fit for your industry?

Look for partners that offer insights into the PLM space and your specific industry.

Like any good PLM system, an implementation partner should be proactive and have an appreciation for moving digital transformation technology forward across all sectors.

Will they provide you with reliable support?

Ensure your PLM partner will offer support at every stage of the implementation process, focusing on the needs of your business with effective solutions that last.

What about the future?

A good PLM implementation partner shouldn't just ensure your solutions and processes work now. Be certain your partner will create a clear, bespoke PLM roadmap that looks years into the future. If they're focused on the here and now without considering the potential twists and turns within your business and industry, you could be in for some nasty surprises.

Related Stories

/Blog AI Mendix

Upgrading to Mendix 11: How Maia and mxcli keep your documentation up to date

Published on Nov 04, 2026
min read
Blog
AI
Mendix

Ask any Mendix developer what they would rather do, build a new feature, or spend half a day documenting the last one, and you already know the answer. Documentation may be far less rewarding than coding, but that is not the only reason it slides to the bottom of the backlog.

Under tight deadlines and pressure to deliver working code, documentation is usually the first thing to go, and what does get written tends to be informal, hard to follow and quickly out of date. The messier the existing docs become, the less point developers see in adding to them, and the gap only grows with every release.

What if documentation could keep itself up to date, scaling with the application and traceable to every change? For many teams, it would change more than daily development. It would change how they look at upgrading to Mendix 11.

 

The hidden cost of undocumented Mendix apps

Many Mendix applications have been running in production for years, some since Mendix 6 or 7, and many of them are business-critical. Upgrading them to Mendix 11 should be an obvious step. Instead, it often feels like a step into the unknown.

When so much of what the app does lives in people's heads, nobody can say exactly what is in it, what will break, or how long it will take to find out, leaving teams faced with three dilemmas:

1. Upgrade now, or keep building on old ground?

Upgrading means facing what years of changes left behind: hidden workarounds, deprecated widgets and old modules nobody remembers. Each major Mendix version also retires older widgets, modules and Java versions, so an upgrade often means replacing parts of the app, not just recompiling it. But staying put means every new feature is built on a foundation that falls further behind, while Mendix 11 capabilities such as Maia stay out of reach.

2. Move fast, or keep it stable?

The business wants new features, more users and new integrations. Without documentation, though, nobody can predict the impact of a change, and without tests or a clear description of how each process should behave, nobody can confirm that everything still works afterwards. Every release becomes a trade-off between speed and stability, and scaling the app gets harder each time.

3. Invest in the upgrade, or in new features?

If you cannot scope an upgrade, you cannot estimate it. Every process has to be checked by hand, and the experienced developers who could answer the questions may have moved on. Faced with an unclear cost and a working app, budgets go to visible features instead. The upgrade keeps getting postponed, and every year it waits, it gets bigger and more expensive.

 

Take the guesswork out of your Mendix 11 upgrade

With Mendix 11, developers finally gain what legacy apps have been missing for years: a clear view of their application and the tools to change it with confidence.

1. Document your Mendix app before you change anything

The first step out of the unknown is a map. Maia, the AI assistant built into Studio Pro, and the Mendix CLI (mxcli), a command-line tool from Mendix Labs that lets coding agents such as Claude Code or Codex work with a Mendix project, can both inspect an application and expose its structure: modules, entities, microflows, pages and the dependencies between them.

An agent can turn that metadata into clear documentation of how the app works, without a developer having to reverse-engineer it by hand. mxcli adds linting and best-practice reports that flag issues and outdated patterns, making technical debt visible before the migration even starts.

2. Move fast without losing control

Once the app is documented, every change becomes easier to judge. mxcli's impact analysis shows which microflows, pages and entities a change will touch before anyone makes it. But more importantly, the documentation no longer falls behind.

After each change, whether a new page, a new attribute or new business logic, the agent updates the documentation as part of the same task, serving two audiences at the same time. Developers and business users read it to understand the app, and AI agents use it to guide the next change or to power an assistant that helps new users find their way.

Because Maia and mxcli work on the Mendix model rather than on thousands of lines of raw code, the output stays easy to review.

3. Turn your Mendix upgrade into an informed decision

When you can see what is in the app, you can scope the upgrade. When you can scope it, you can estimate it, and the budget discussion becomes a business decision instead of a leap of faith.

The platform you upgrade to matters too. With Mendix 11.12, the first Long-Term Support (LTS) release of Mendix 11, teams get a stable foundation for long-term maintainability, security and scalability, and the AI-assisted tooling that comes with it is ready for business-critical production apps. Maia Plan turns a description into epics and user stories, and Maia Make starts implementing them, so the investment keeps paying off long after the migration.

4. Lay the foundation for the agentic enterprise

An application that documents itself, is also ready for what comes next. This is exactly what companies get with Mendix in the context of enterprise orchestration.

As part of Siemens Intelligence Center X (ICX), Siemens' industrial AI orchestration software, Mendix, together with Maia and mxcli, plays an important role in making applications understandable to agents and people alike. With Mendix 11.12, an app can act as an MCP server that exposes its logic, data and workflows to other AI tools, and Mendix Workflows can orchestrate agents and people inside a governed, auditable trust layer, creating the foundation for the hybrid workforce that brings people and AI agents together.

 

See mxcli in action: one prompt, feature and documentation done

In a recent demo, Erick Souza, Sr. Mendix Developer & Mendix MVP at CLEVR, showed how a single prompt can build a new feature and update the documentation in the same run. No one has to sit down and write it afterwards. Here is what he showed, if you don't have time to watch the full video.

Could a developer make these changes by hand? Of course, but it would take longer. Here it happens automatically, every time.

 

Upgrade to Mendix 11 with confidence, starting with your documentation

Download mxcli from the Mendix Labs GitHub page, drop it into your project folder and run mxcli init. Within minutes, your project is ready for the coding agent of your choice, whether that is Claude Code, Codex or another tool.

From there, the first prompt is the most valuable one: ask the agent to document what is already in your app, then make updating the documentation part of every prompt, and your team's knowledge grows with the application instead of falling behind it. One word of caution: mxcli is still alpha software and evolving quickly, so always work on a copy of your project or under version control, and review the agent's changes the way you would review a colleague's work.

The result is a foundation you can build on. An application your team understands, an upgrade you can plan, and an architecture ready for the agentic enterprise.

Curious what this could look like for your applications? Contact CLEVR to request a demo.

November 4, 2026 5:08 PM
/Blog AI Intelligence Center X

From data points to decisions: How ontologies and knowledge graphs ground the agentic enterprise

Published on Nov 06, 2026
min read
Blog
AI
Intelligence Center X

Every agent in your enterprise must answer one question before it acts: what am I actually looking at? A language model can read a record from Teamcenter, SAP and your MES. On its own, it cannot know that the “part” in all three is the same physical thing, that this part sits in an assembly under an open engineering change, and that last week’s warranty claim points at exactly that assembly.

Your people know this from years of experience. Agents need it written down. That written-down understanding is an ontology, and once it is filled with live data it becomes a knowledge graph. Together they form the least glamorous and most decisive layer of the agentic enterprise: the one that lets people and agents work from the same reality.

This blog explains what ontologies and knowledge graphs are, how they work, where they sit in your architecture, what the predefined industrial ontologies in Siemens Intelligence Center X bring to manufacturers, and what it takes to build one yourself.
‍

Short on time? Here’s a brief overview

  • Agents fail on context, not on models. Gartner expects 60% of AI projects without AI-ready data to be abandoned. The difference between an agent that creates value and one that creates incidents is whether it understands what your data means.
  • An ontology is the rulebook; a knowledge graph is the map drawn by that rulebook. The graph links every part, change and claim across your systems, so agents can follow the path from question to answer. Together they give people and agents the same definitions, the same rules and the same audit trail, the basis of a hybrid workforce you can trust.
  • Keep the core clean and layer context on top. Your systems of record stay untouched at the base. The ontology and knowledge graph sit directly above them as a shared context layer, with orchestration and AI added on top where they make sense.
  • Not every ontology layer is built for agents. Agents ask questions nobody anticipated, across many systems at once. Graph Studio in Siemens Intelligence Center X is a scalable graph platform designed to answer those at enterprise scale, without performance limits.
  • Build on proven industry knowledge. You don’t have to model manufacturing from scratch. Predefined industrial ontologies and open standards already describe the digital thread, from engineering BOM to warranty claim. Your domain experts make them your own, matching records across systems and governing how the model evolves. Start with one cross-system workflow and let every next use case extend it.
    ‍

Why agents stall without context

The model is rarely the bottleneck. The context it reasons against is.

Gartner predicts that through 2026, organisations will abandon 60% of AI projects that are not supported by AI-ready data, and 63% of organisations either lack the right data management practices for AI or are unsure whether they have them. For agents specifically, Gartner expects more than 40% of agentic AI projects to be cancelled by the end of 2027, citing escalating costs, unclear business value and inadequate risk controls.

The clearest evidence of what context does comes from a benchmark by data.world researchers. They asked GPT-4 business questions over an enterprise SQL database from the insurance domain. Queried directly, the model answered 16.7% of questions correctly. Queried over a knowledge graph built on an ontology of the same data, accuracy rose to 54.2%. On the hardest questions over the most complex schemas, direct SQL scored 0%, while the knowledge graph route reached 36–39%. A follow-up study used the ontology itself to check and repair the model’s queries and pushed accuracy to 72%.

Same model, same data. The only difference was whether the meaning of the data was made explicit. For an agent that does not just answer but acts, that difference decides whether it creates value or creates incidents.

Why 72% is a starting point, not a verdict

No enterprise should act on 72%, and no well-designed enterprise system asks it to. The benchmark measures a language model writing its own queries against open questions, without guardrails. In production, three things change the picture.

First, the ontology acts as a check, not just a guide. Queries that do not fit the model of the business are caught and repaired, or the system declines to answer. A known “I don’t know” is safe. A confident wrong answer is not.

Second, the required accuracy follows the consequence. Exploring a question can tolerate uncertainty. Approving a recall cannot. High-stakes decisions run on engineered rules and validated queries, with a human approval step, while open exploration stays a support tool.

Third, every answer is traceable. Which data was used, which query ran and who approved the action is logged, so an error is visible and correctable instead of hidden.

The takeaway from this research is not that AI using ontologies is 72% reliable. It is that grounding AI in context is the most effective way to make it more reliable, and to make it honest about where its knowledge ends.
‍

The rulebook and the map

An ontology is the rulebook; a knowledge graph is the map drawn by that rulebook. The two terms are often used interchangeably, but they do different jobs.

The ontology is a formal vocabulary for your business. It defines the types of things that exist (a pump, a production line, an engineering change, a supplier), the properties they have, how they may relate to each other, and the rules that apply. A pump can be installed in a line. An engineering change affects one or more parts. A part cannot be released without an approved design. Without an ontology you have a picture of dots and lines. With one, the system knows what each dot is and can reason about it.

The knowledge graph is that vocabulary applied to your real data: Pump P-101, Line 3, change request ECR-2291, and the millions of connections between them, refreshed from your systems of record.

Aspect Ontology Knowledge graph
What it is The schema of meaning: types, relationships, rules The instances: your actual assets, parts, orders and their links
Example “A Part is affected by an Engineering Change” “Part 4471-A is affected by ECR-2291”
Size Hundreds to a few thousand concepts Millions to billions of facts
Rate of change Slow, versioned, deliberately governed Continuous, synchronised with source systems
Owned by Domain experts and knowledge engineers Data pipelines, agents and the platform

This split matters for the agentic enterprise. The ontology is where people encode what the business means. The graph is where agents and applications read and act on it. Get the ontology right and every new agent inherits that understanding for free.
‍

How a knowledge graph works

A knowledge graph stores every fact as a triple: subject, relationship, object. Chain enough triples together and you get a map that people and agents can walk, rather than a warehouse they have to query with a dozen painful joins.

Triples. “Pump P-101 is installed in Line 3.” “Line 3 produces Product X.” “Product X contains Assembly 4471.” “Assembly 4471 has requirement max 80 °C.” Each thing is a node, each relationship an edge. In the W3C standards most enterprise graphs use, these are RDF triples, queried with SPARQL.

Traversal. A question like “which open warranty claims involve parts touched by last month’s engineering changes?” crosses PLM, ERP and service data. In a relational world that is a project. In a graph it is a path: claim → serial number → as-built configuration → part → change. Agents are good at following paths, as long as the paths exist and mean something.

The ontology defines the path; the graph holds your data
Which engineering change touches this warranty claim? One path across service, ERP, MES and PLM data.
Ontology: the rulebook
Knowledge graph: the map
Warranty claim
Claim from last weekservice system
reported on
Product unit
Serial number of the unitERP
built as
As-built configuration
As-built recordMES
contains
Part
Part 4471-APLM and ERP, resolved as one
affects
Engineering change
ECR-2291PLM

Ontology vs Knowledge graph · one path from claim to change

The left column is written once, by people. The right column is filled continuously from your systems. An agent that understands the left can find the right for any claim, in any plant.

Inference. Because the ontology carries rules, the graph can derive facts nobody entered. If every assembly inherits the compliance requirements of its parts, a new regulation on one material instantly flags every product that contains it.

Entity resolution. The hardest, least visible part. “Pump_P101” in the maintenance system, “P-101” in the MES and “asset 4471-A” in ERP must be recognised as one object, across millions of records. Get this wrong and agents confidently reason about three pumps that are really one.

Materialised or virtual. A graph can copy data into a graph store or leave it where it is and map it on the fly. Bosch’s plant in Salzgitter used the virtual approach: an ontology for surface-mount electronics manufacturing, mapped onto existing databases. Machine failures recorded differently across machines were reconciled into a single ontology class, so engineers could finally compare like with like.

Grounding generative AI. The same structure makes retrieval for language models sharper. LinkedIn built a graph from historical customer-service tickets instead of treating them as loose text. In production, the system cut median per-issue resolution time by 28.6%. The lesson carries straight to a service or quality organisation in manufacturing: structure beats a bigger pile of documents.
‍

Where it sits: the foundation layer

In the CLEVR Digital Renovation Principle, the core systems form Layer 1: ERP, PLM, CRM, MES and in-house applications, the systems of record that stay clean and stable. On top of that core, the ontology and knowledge graph form Layer 2: the semantic context layer that every layer above it reasons against. Orchestration and intelligence live in Layer 3. People meet agents in Layer 4. The graph does not replace any system of record; it makes them understandable to each other. Together, the layers put the CLEVR principle into practice: keep the core clean, layer a smart orchestration platform on top of it, and amplify with AI where it makes sense.

Keep the core clean, layer with a smart orchestration platform on top and amplify with AI where it makes sense.
CLEVR Digital Renovation layers mapped to Siemens Intelligence Center X
Layer 4 · Human-agent experience Approvals, exceptions and oversight: where people meet agents
Layer 3 · Orchestration and intelligence Mendix workflows, apps and agents · AI Studio models, applied where AI adds value
Layer 2 · Semantic context Graph Studio: connects the systems below, gives meaning to every layer above
Layer 1 · Core systems Systems of record: PLM, ERP, MES, CRM and in-house apps, kept clean and stable
Enterprise trust layer
Guardrails
Traceability
Audit

Four layers · a clean core at the base, semantic context on top of it

Every agent and app in the upper layers draws its meaning from the same foundation, under the same trust layer.

Siemens Intelligence Center X is built on the same logic, in four steps:

  1. Enterprise data from heterogeneous sources and data platforms.
  2. Context: an ontology layer spanning the enterprise that gives every data point its meaning, delivered by Graph Studio.
  3. Intelligence: AI and ML models in AI Studio that learn from semantically enriched data, refreshed in near real time.
  4. Action: applications, workflows and agents built and orchestrated on Mendix, all rooted in the same context.

An enterprise trust layer runs across all four, with guardrails, traceability and auditability for both people and agents. Graph Studio’s engine is designed to handle tens to hundreds of billions of RDF triples, in memory, on disk or virtualised, and exposes the graph to agents through MCP so they can explore and query across domains.

The wider market is converging on the same idea. Microsoft made Fabric IQ, its ontology-based context layer for agents, generally available at Build 2026, and is opening the ontology to agents from other vendors through MCP. Databricks and Google have moved their semantic layers in the same direction, and Palantir has built Foundry around its Ontology for years. The question is no longer whether you need a semantic layer for agents. It is which one understands your domain, and whether it connects to the platforms you already run. Intelligence Center X connects to Microsoft Fabric, Snowflake, Databricks and Palantir rather than competing with them, and it differentiates on two things: the depth of industrial meaning it brings on day one, and a graph architecture built for the questions nobody anticipated.
‍

Different approaches to the ontology layer

Not every ontology layer is built the same way. Look under the hood and the market splits into four architectural approaches, each optimised for a different job.

Approach Examples Optimised for How the ontology is built
Metadata layer: a semantic model over tables Databricks, Snowflake, Microsoft Fabric, AWS Governed SQL analytics over tables you already have Semantic views authored over existing warehouse tables
Object layer over a data lake Palantir Foundry Operational decisions on curated objects and actions Object types and pipelines defined per source
Transactional graph database Neo4j, Amazon Neptune, Stardog Traversal and pattern matching around a known starting point Labels and relationships modelled, then loaded
Scalable graph platform Siemens Graph Studio (Graph Lakehouse) The unanticipated multi-hop question, at enterprise scale Generated from the source and linked automatically

The first three are strong at what they were designed for. A metadata layer makes dashboards and SQL analytics consistent. An object layer works well when the decisions are known in advance and the objects can be curated for them. A transactional graph database is fast when a question starts from a known node, such as one customer or one asset.

Agents do not work that way. A hybrid workforce asks questions nobody anticipated when the data model was designed: which suppliers, parts, plants and customers are exposed if this material is restricted, across every product line? Such a question has no single starting point and crosses many hops and domains at once. On a metadata or object layer, it means building new models or new pipelines first. On a transactional graph database, it means traversing outward one node at a time, which slows down as the graph and the question grow.

Of the four approaches, only a scalable graph platform is designed for that question. Graph Studio, the context layer of Intelligence Center X, runs on a massively parallel engine that distributes the graph across the cluster, so every core works on every query. Its ontology is generated from the source systems and linked automatically, rather than authored by hand per source. That is the difference between agents that answer the questions you planned for and an agentic enterprise in which people and agents can ask any question, without running into performance limits.

This does not make the other approaches obsolete. Intelligence Center X connects to the data platforms you already run, and adds the context layer that lets agents reason across all of them.

‍

Predefined industrial ontologies: not starting from a blank page

A predefined industrial ontology is a ready-made model of what things mean in manufacturing, so you extend it rather than invent it. This is where Intelligence Center X sets itself apart from generic semantic layers.

Siemens ships Intelligence Center X with industrial ontologies that encode industry reasoning across the product lifecycle. Out of the box, they cover the core objects of the digital thread: design, part, engineering BOM, engineering change, bill of process, service BOM, warranty claim, inventory and IoT time series. These are connected live to systems of record (PLM, ERP, CRM and in-house applications) and to data platforms such as Microsoft Fabric, Snowflake, Databricks and Palantir. Siemens describes this as data in context that is “live, secure, configured, out of the box”.

Why this matters is easiest to see in the BOM. An engineering BOM describes what the product is. A manufacturing BOM describes how it is made. A service BOM describes how it is maintained. In most companies these live in different systems, owned by different teams, with different identifiers. An ontology that already knows how an EBOM item relates to its MBOM counterpart, which engineering change affects both, and which service parts follow from them gives agents the digital thread as a navigable structure. It also reflects decades of Siemens experience with Teamcenter, Opcenter and the wider Xcelerator portfolio, knowledge that is expensive to recreate from scratch.

Siemens is not alone in this field. Several open initiatives define manufacturing semantics, and a good ontology strategy knows where they fit:

Source What it covers Maintained by Where it helps
Siemens Intelligence Center X industrial ontologies Digital thread objects: EBOM, MBOM, change, bill of process, service BOM, warranty, inventory, IoT Siemens Digital Industries Software Fast start for Siemens-centric product lifecycles, live-connected to source systems
IOF Core Ontology Common manufacturing concepts (manufacturing process, assembly, material product, agent roles), plus domain modules for production planning, maintenance and biopharma Industrial Ontologies Foundry (OAGi), with NIST; MIT licence Shared vocabulary across companies and suppliers, built on the ISO/IEC 21838-2 top-level ontology
Asset Administration Shell (IEC 63278) Digital twin of an asset, with standard submodels such as Digital Nameplate, Technical Data and Handover Documentation Industrial Digital Twin Association (IDTA) Exchanging asset data across the supply chain, Manufacturing-X and Digital Product Passport

These are complementary rather than competing. The IOF Core is designed so that domain ontologies stay consistent with each other (IOF). The Asset Administration Shell structures asset information in submodels whose templates the IDTA publishes (all-about-industries). A predefined platform ontology gets you to production fastest; mapping it to open standards keeps you interoperable with suppliers and customers.

One honest caveat: no predefined ontology knows your company. Your product families, naming conventions, plant-specific processes and in-house systems still have to be modelled. Predefined ontologies shift the work from inventing to extending, which is the difference between months and weeks.
‍

What it takes to build an ontology

Ontology design is a craft, not a configuration screen. It succeeds when it starts from decisions and is built with the people who make them. In our experience, it follows seven steps:

  1. Start from the decisions, not the data. Write down the questions agents and people must be able to answer: “Which customers are affected if supplier X fails?” “Which products are touched by this change?” Knowledge engineers call these competency questions. They define the scope and become the test set.
  2. Reuse before you invent. Start from a predefined industrial ontology and open standards such as the IOF Core or the Asset Administration Shell. Only model what makes your company different.
  3. Model with domain experts in the room. Engineers, planners and service leads know what a “release” or a “variant” really means. The Bosch integration work found that domain expert involvement was essential for both the ontology and its mappings.
  4. Map your sources. Connect fields in Teamcenter, SAP, the MES and your in-house apps to ontology concepts. Decide per source whether data is copied into the graph or queried virtually.
  5. Resolve entities. Decide which records describe the same real-world object, and keep those decisions explainable.
  6. Validate against reality. Run the competency questions. Check the ontology for logical consistency with a reasoner. Let agents query it and look at where they take wrong paths.
  7. Govern and evolve. Version the ontology, assign owners per domain, and keep the graph in sync with its sources. A graph that falls behind its source systems is yesterday’s data, and an agent acting on it is acting on the past.

This used to be the reason knowledge graph projects took a year before delivering value. Automation is changing that. Siemens positions automated ontology creation from source systems as a way to remove months-long ontology development and uses MCP-connected agents to help create and maintain the knowledge graph itself. Research points the same way: at Bosch, automatically reshaping a domain ontology into a leaner graph schema made knowledge graph generation for welding data seven to eight times faster.

Automation speeds up steps 2 to 5. It does not replace steps 1, 3 and 7. Deciding what your business means, and who is accountable for that meaning, stays human work.
‍

One ontology for a hybrid workforce

A hybrid workforce only works when people and agents share the same definitions, the same rules and the same boundaries. The ontology is where all three live.

Same definitions. When a planner and a planning agent both mean the same thing by “available stock” or “released part”, hand-offs stop producing errors. Without that, every agent quietly carries its own interpretation in its prompt, and you have rebuilt the silos you were trying to escape.

Same rules. Policies expressed against business concepts rather than database tables apply to everyone. Intelligence Center X adds policy-driven access control to the knowledge graph, based on attributes of both the user and the data, so the same rule governs what a human user and an agent may see. A supplier-quality agent sees what a supplier-quality engineer may see, and nothing more.

Same trail. Because every action references concepts in the graph, the trust layer can log not just what an agent did but what it was reasoning about. That is what makes decisions auditable, and autonomy defensible.

This also creates new work for people. In our Digital Renovation whitepaper series we described roles like the Agent Supervisor and the Eval Owner. Add one more: the Ontology Steward, the domain expert who owns what a concept means and approves changes to it. Researchers have called this profile the knowledge scientist. Whatever you call it, organisations that skip it end up with a graph nobody trusts.
‍

Where to start: one thread, not the whole enterprise

You do not need an enterprise-wide ontology before your first agent goes live. You need one valuable thread through your systems, modelled well, and a foundation that grows with every next use case.

Pick a workflow that crosses systems and hurts today. Quality is a good candidate. Siemens illustrates a typical warranty-to-engineering loop in which complaint triage, investigation and root cause analysis take two to four weeks, and resolution around eight. That loop touches warranty claims, service BOMs, as-built configurations, parts and engineering changes: exactly the concepts the predefined industrial ontologies in Intelligence Center X already cover.

From there, the CLEVR approach is the one we apply to every agentic programme: think big, start small, scale fast. Think big by designing the ontology as the shared foundation for your whole hybrid workforce. Start small with one thread, extended from a predefined ontology rather than drawn from scratch. Scale fast by letting every new agent, app and workflow reuse what the first one taught the graph.

Agents will keep getting smarter. Whether they get smarter about your business depends on the layer underneath them.

November 6, 2026 1:00 PM
/Blog Aerospace & Defense Teamcenter

Digital shipbuilding: the shipyard bottleneck isn't steel, it's data

Published on Nov 02, 2026
min read
Blog
Aerospace & Defense
Teamcenter

Even though EU defense spending reached €418 billion in 2025, a 47% rise since 2022, and is projected to climb to €454 billion in 2026, the European Defence Agency (EDA) warns that, under current national plans, Europe's shortfalls in naval forces are unlikely to be closed by 2030.

For naval shipbuilding, the pressure to deliver more vessels, faster, has never been higher, and the constraint isn't simply dry docks, steel or skilled welders. It's fragmentation, both between countries and inside organizations. Ask how a yard keeps thousands of parts, drawings and revisions in sync across a naval program, and the answer varies widely: modern shipbuilding software in one yard, a legacy PDM system and a stack of spreadsheets in the next. Even within a single yard, engineering data is often split across systems, with an ERP that doesn't talk to either.

With a stronger push for European sovereignty and growing attention to sustainability, that fragmentation is becoming harder to ignore, and digital shipbuilding is fast becoming a necessity rather than an option.

 

6 data challenges holding back naval shipbuilding

Steel decides how much a yard can build. Data decides how much of it is built right the first time. In naval shipbuilding, programs run for years, requirements shift mid-build and every decision has to be traceable from day 1. In that environment, fragmented data doesn't stay contained. It cascades through every stage of a program, adding rework, delays, and stretching lead times.

1. Outdated designs on the production floor

Naval programs rarely have the luxury of a finished design. To meet delivery dates, yards cut steel for early blocks while later sections are still being engineered, and the pressure to deliver faster pushes that overlap even further.

When engineering data is split between a legacy PDM tool and spreadsheets, nobody can say with certainty which revision production is working from. In practice, that might mean a block being outfitted with a pipe route that engineering changed two weeks earlier.

2. Design changes discovered too late

A naval vessel is a tightly coupled system, and its requirements often evolve during the build as systems and mission profiles change. Swap one piece of equipment and the change can ripple into structure, weight, cabling, ventilation and the bill of materials. When that information sits in disconnected files, the ripples surface late, on the production floor, where they cost the most to fix.

3. Mismatched bills of materials

Engineering defines what the ship is made of. ERP buys, plans and schedules it. When the two aren't connected, people bridge the gap by hand: re-typing part data, reconciling lists, chasing discrepancies. With long-lead naval equipment, an order placed against an outdated revision can mean months of delay, not days.

4. Person-dependent approvals

In many yards, releasing a drawing or approving a part depends on a few experienced people who know how it works. With a generation of engineers retiring, that is a continuity risk. And defense customers expect every decision to be traceable, which an approval trail spread across email and spreadsheets cannot reliably provide.

5. Hidden costs and carbon footprint

Most of a vessel's cost is locked in during design, yet it is often only calculated later. When one supplier delivers late, the impact ripples through the build sequence: outfitting stalls, blocks wait in the yard, and work is resequenced or done out of order at a higher cost. On the shop floor, rework, extra labor hours and expedited parts add up without ever being traced back to the change that caused them.

By the time the real number surfaces, there is little room left to steer.

6. No transparency on program status

Defense customers expect regular, accurate progress reporting. When data is scattered, simple questions become hard: What has been released? What is waiting for approval? Which changes are still open?

And reports compiled by hand from contradictory sources rarely offer the traceable, accurate picture engineers can stand behind.

 

The digital thread: how digital shipbuilding connects design, production and cost

A digital thread is a continuous flow of information that connects every stage of a vessel's lifecycle, from the first design decision to the delivered ship. Where a digital twin is the virtual model of the vessel, the digital thread is what keeps that model, and every team working on it, supplied with accurate, traceable data.

By connecting design, engineering, production, cost and performance data this way, shipbuilders can create a living digital twin of each vessel, automate the handovers that slow yards down, and take control of every stage of the build.

1. Design and engineering in one model

Every change made at the drawing board reaches the parts list and the production drawings automatically. Integrated CAD software makes it possible for designers and engineers to work on the same model, and design data flows into engineering without exporting, redrawing or re-entering it. The result is fewer drawing errors and faster design iterations.

2. One source of truth for every part and revision

Every part, drawing and revision has a single home, with a clear status and history. Shipbuilding PLM software, built on product lifecycle management, assesses each change against the full product structure before release, and routes approvals through standardized workflows instead of email chains. Know-how stays in the system when experienced engineers leave, and defense customers get a complete record of who approved what, and when.

3. Engineering and business in sync

Purchasing, planning and production work from the same bill of materials, so long-lead equipment is ordered against the right revision. By connecting PLM to ERP and other business systems with tailored low code applications, released engineering data flows automatically into purchasing and planning, and nobody re-types part lists by hand. That means less reconciliation work and fewer costly ordering mistakes.

4. A shop floor that sees what engineering sees

The shop floor always builds from the latest released design, and what happens there flows back to engineering. A manufacturing execution system (MES) connected to the digital thread captures progress, rework and extra hours as they happen, and provides digital work instructions and a complete as-built record of every vessel. When a supplier is late, plans can be adjusted with a clear view of what that means for the rest of the build.

5. Cost and carbon decided while the design is still open

Teams see what a design change or supplier delay will cost before it reaches the budget. Product cost management software (PCM) calculates cost and carbon footprint while the design is still open, including complex, non-linear pricing models from suppliers. It also gives buyers a fact-based position in supplier negotiations and a reliable basis for quotes and tenders.

6. Live insight instead of manual reports

Program managers, engineers and customers see real progress as it happens, backed by data they can trace and trust. An industrial IoT and analytics layer brings data from engineering, the shop floor and the business together in tailored dashboards, with live insight into equipment performance and early warning when production drifts off plan.

7. Intelligence that moves the program forward

Intelligent process orchestration flags risks early, anticipates bottlenecks and coordinates the next step across teams and systems, so issues are resolved before they turn into delays. And because it is built on the same structured, traceable data, AI stays governed. Every action is logged, access follows the same roles and permissions as the rest of the thread, and the orchestration is tailored to each yard's own processes. That makes it compliant by design and able to scale from a single workflow to an entire program.

 

From spreadsheets to a digital shipyard: a CLEVR digital transformation example

With more than 30 years of experience in digital transformation, CLEVR supports PLM, manufacturing operations and application programs for multiple aerospace, defense and naval manufacturers across Europe. One of them, a leading European shipyard building naval and government vessels, faced exactly the problems described in this article.

The problem: No digital thread, no traceability

Keeping parts, drawings and specifications in sync took constant manual effort, releases and approvals depended on the know-how of individual engineers, and reporting was limited. With engineering data spread across a legacy PDM system and disconnected spreadsheets, and no link to ERP, our client needed a system that understood how a shipyard actually works.

The process: Implementing shipbuilding product lifecycle management software

Rather than treating shipyard digital transformation as a big-bang replacement, the CLEVR consultants introduced Siemens Teamcenter, an industry-leading PLM platform, in clearly defined phases. First, they implemented Teamcenter on the yard's own servers, allowing for full control over its engineering data from day one, and then, they configured it around real engineering practice. This meant mapping out how custom types and properties should describe parts, drawings and documents the way the yard's engineers actually work with them, establishing roles and access rights that define who can create, change and release what, and standardizing release workflows so every drawing and part follows a defined route from draft to approval.

The result: Value that is built to scale

That foundation is now live: every part, drawing and revision has one home and a traceable history, and approvals no longer depend on who happens to know the process. More importantly, it gives the shipyard a base to build on. Advanced workflows, reporting, CAD integration and the connection to ERP are already being rolled out, each one adding automation on top of data that is structured and trusted. Once the thread is connected end to end, intelligence can be added to anticipate issues, orchestrate work across teams and help the yard grow faster as its programs scale.

 

CLEVR: Your go-to partner for shipyard digital transformation

This is the true power of the CLEVR methodology.

Rather than forcing a standard template, our consultants take the time to understand how each shipyard engineers its vessels, and configure the system around that. We work in phases, so every step delivers something usable before the next one begins, and we stay alongside our customers' teams from day one. We are not rebuilding old integrations as they were, we design for fewer manual steps from the start. That is what turns a software implementation into a change engineers feel in their day-to-day work.

Want to see how a connected digital thread can help your shipyard turn growing demand into vessels delivered on time? Contact CLEVR for a consultation.

November 2, 2026 9:06 AM

Frequently Asked Questions

1

What does PLM stand for?

PLM stands for Product Lifecycle Management.

2

What are the steps in the PLM process?

The PLM process is divided into five main stages: Conception, Design and Engineering, Manufacturing, Commissioning, and Decommissioning.

3

What is a PLM strategy?

A PLM strategy is a strategic approach to developing, managing, and improving products from conception to disposal. It creates a framework that blends existing procedures, individual expertise, and technology to enhance product quality, reduce costs, and accelerate time to market.

4

What is the difference between PLM and PDM?

PDM (Product Data Management) is a key component within the broader PLM system. While PDM focuses specifically on centralizing and managing product-related data (such as version control and access permissions), PLM is the overarching system that manages the entire product lifecycle and all associated processes.

5

What is the difference between ALM and PLM?

The primary difference lies in the nature of the product being managed: PLM is designed for the development of physical products and manufacturing processes, handling everything from initial conception and manufacturing specifications to decommissioning. In contrast, ALM (Application Lifecycle Management) is focused on the development of software applications and digital systems.

While both share core management principles, their applications differ significantly. For example, PLM stages include complex physical requirements like prototyping, mass-production scaling, and environmental decommissioning, whereas ALM focuses on code iterations and software releases. Consequently, PLM requires its own specialized toolset (like Siemens Teamcenter), though agile ALM tools and low-code platforms can be adapted to extend and optimize these PLM processes.

Contact us

Want to know how our solutions, products, and services can accelerate your digital transformation? 

Want to know how our solutions, products, and services can accelerate your digital transformation?